HK1064849A - Method and apparatus for reducing latency in waking up a group of dormant communication devices - Google Patents

Description

Method and apparatus for reducing latency in waking up a group of dormant communications

Reference to related applications

This application claims benefit of U.S. provisional patent application serial No. 60/291454 filed on 5/15/2001, which is hereby incorporated by reference in its entirety.

FIELD

The present invention relates to point-to-multipoint communication systems, and more particularly to a method and apparatus for reducing latency in waking up a group of dormant communications.

Background

Wireless service classes for fast, efficient, one-to-one or one-to-many (group) communication have existed in a variety of forms for many years. Typically, these services are half-duplex, in which the user presses the "push-to-talk" (PTT) button of his phone/radio to start speech. Pressing a button or in some embodiments automatically turning on the radio or in a medium system where communication occurs through some type of server, indicates the user's request for "talk". The user can talk for typically a few seconds if authorized to speak or allowed by the interlocutor, and other speakers can request to speak after he releases his PTT { XE "PTT" }. Communication is typically from one speaker to a group of listeners, but may be one-to-one. This service is typically used in applications where one person, the "dispatcher," needs to communicate with a group of people, such as field service personnel or taxi drivers, which is the source of the "dispatch" name of the service.

Recently, there have been similar services on the internet and are generally referred to as "voice chat". These services are typically implemented as personal computer { XE "PC" } applications that send vocoder frames to a central group chat server within Internet (IP) { XE "IP" } packets, i.e., Voice over IP (VoIP) services, or possibly from client to client within a peer-to-peer service.

A key feature of these services is the rapidity and spontaneity, typically initiated simply by pressing the PTT { XE "PTT" } button, without the need for a typical dialing and ringing sequence. Communication within this type of service is typically short, with individual speech typically being concentrated in a few seconds, while "talking" may typically last a minute or less.

The time delay between the user requesting the talk and when he receives a positive or negative acknowledgement from the server that he has the floor and may start speaking, called PTT { XE "PTT" } latency, is a key parameter of the half-duplex group communication system. As described above, the scheduling system is most important for short and fast conversations, which reduces service efficiency if the PTT latency is too long.

Existing group communication infrastructures provide limited opportunities to greatly reduce PTT latency, i.e., the actual PTT latency may not be reduced below the time required to re-establish traffic channels within a dormant packet data session. In addition, the speaker and listener channels are concatenated because the only available mechanism to wake up the dormant group is to wait for the speaker's traffic channel to be re-established to signal the server. Currently, there is no mechanism to be able to send mobile station user-oriented signaling data on any other channel than the traffic channel — this limitation requires that the traffic channel be re-established before any communication between the client and server can begin.

There is therefore a need for a mechanism to reduce the PTT latency experienced by the speaker and to reduce the total time, customer battery life, and other resources needed to re-establish traffic channels to enable mobile stations to participate without adversely affecting system capacity.

Summary of the invention

The disclosed embodiments provide novel and improved methods and apparatus for reducing sleep-wake latency in a group communication network. In one aspect of the invention, a method of reducing latency in a group communication network comprises the steps of: receiving a sleep-wake trigger directed to a target listener mobile station; triggering the target listener mobile station to re-establish the traffic channel; and storing the sleep-wake trigger for later transmission to the target listener mobile station. In one aspect, the method further includes transmitting the stored sleep-wake trigger to the target mobile station once the target mobile station has re-established the traffic channel.

In one aspect, an apparatus for reducing latency in a group communication network includes a memory unit, a receiver, a transmitter, and a processor communicatively coupled to the memory unit, the receiver, and the transmitter. The processor can receive a sleep-wake trigger directed to the listener mobile station, trigger the target listener mobile station to re-establish a traffic channel, and store the sleep-wake trigger for later transmission to the target listener mobile station.

Brief description of the drawings

The features, nature, and advantages of the present invention will become more apparent from the detailed description set forth below when taken in conjunction with the drawings in which like reference characters identify correspondingly throughout and wherein:

FIG. 1 illustrates a group communication system;

FIG. 2 illustrates how several communication devices interact with a communication manager;

fig. 3 illustrates call signaling details of a floor control request procedure according to an embodiment;

FIG. 4 illustrates network initiated sleep-wake call signaling details according to one embodiment;

FIG. 5 illustrates buffering media at a communication manager, according to an embodiment;

FIG. 6 illustrates buffering media at a client in accordance with one embodiment; and

fig. 7 illustrates an example radio link mode in accordance with an embodiment.

Detailed description of the invention

Before one embodiment of the invention is explained in detail, it is to be understood that the invention is not limited in its application to the details of construction and the arrangement of components set forth in the following description or illustrated in the following drawings. The invention is capable of other embodiments and of being practiced or of being carried out in various ways. Also, it is to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting.

Fig. 1 illustrates an exemplary functional block diagram of a group communication system 100. Group communication system 100 is also referred to as a push-to-talk system, a Network Broadcast Service (NBS), a dispatch system, or a point-to-multipoint communication system. Within the NBS 100, a group of communication device users, individually referred to as net members, communicate with each other using communication devices assigned to each net member. The term "net" refers to a group of authorized users of communication devices that are able to communicate with each other.

In one embodiment, the central database may contain information identifying the members of each particular network. There may be more than one network operating within the same communication system. For example, a first net may be defined as having ten members and a second net may be defined as having 20 members. Ten members of a first net may communicate with each other but may not be able to communicate with members of a second net. In another embodiment, members of different networks may be able to monitor communications between members of more than one network, but may only be able to send information to members within their own network.

The network may operate over an existing communication system without requiring major changes to the existing infrastructure. Thus, controllers and users within a network may operate within any system that transmits and receives packet information using Internet Protocol (IP), such as Code Division Multiple Access (CDMA) systems, Time Division Multiple Access (TDMA) systems, global system for mobile communications (GSM) systems, satellite communication systems such as Globalstar^TMOr Iridium^TMOr a variety of other systems.

The network members can communicate with each other using assigned communication devices, shown as Communication Devices (CDs)102, 104, 106, and 108. The CDs 102, 104, 106 and 108 may be wired or wireless communication devices such as land-based wireless telephones, wired telephones with push-to-talk functionality, satellite telephones with push-to-talk functionality, wireless video cameras, still cameras, audio devices such as music recorders or players, laptop or desktop computers, paging devices, or any combination thereof. For example, CD 102 may comprise a wireless terrestrial telephone with a video camera and playback. In addition, each CD may be capable of sending and receiving information in either a secure mode or a non-secure (open) mode. In the following discussion, a single CD refers to a wireless push-to-talk telephone. It is to be understood, however, that reference to a CD is not intended to be limiting and may include other communication devices capable of transmitting and receiving packet information in accordance with the Internet Protocol (IP).

In the NBS system 200 of fig. 2, the transmission preferences generally allow a single user to transmit information to other net members at a given time. The transmission priority generally determines whether the requesting network member can obtain a transmission priority based on whether the current transmission priority is assigned to another network member. The process of granting and denying transmission requests is referred to as arbitration. The arbitration scheme may weigh factors such as the priority assigned to each CD, the number of unsuccessful attempts to obtain a transmission priority, the length of time that a network member retains a transmission priority, or other factors in determining whether to grant a requesting network member a transmission priority.

To participate in the NBS system 100, the CDs 102, 104, 106, and 108 may each be able to request a transmission priority from a controller or Communication Manager (CM) 110. The CM 110 may manage the real-time and administrative operations of the web. A CM is any type of computer device with at least one processor and memory. In one embodiment, CM is Sun workstation Netra T1^TM。

The CM 110 may operate remotely by either the communication system service provider, a web member, or both, assuming approval is provided by the service provider. The CM 110 may receive the web definition through an external administrative interface. Web members may request administrative actions through their service providers or request administrative web functions through defined systems such as the Security Manager (SM)112 operated by members conforming to the CM administrative interface. The CM 110 may authenticate a party attempting to establish or modify a web.

SM112 may implement key management, user authentication, and related tasks to support security nets. A single group communication system may interact with one or more SMs 112. The SM112 may not be involved in real-time control of network entry, including network activation or PTT arbitration. The SM112 may have administrative capabilities compatible with the CM 110 to automate administrative functions. The SM112 may also be able to act as a data endpoint to participate in the network, broadcast network keys, or simply monitor network traffic.

In one embodiment, the means for requesting transmission priority from the CM includes pressing a push-to-talk (PTT) button or switch. When a user in NBS 100 desires to send information to other net members, the user may press a push-to-talk switch located on his or her CD, sending a floor control request to get a transmission priority from CM 110. If no other net member is currently assigned a transmission priority, the requesting user may be granted a transmission priority and the user may be notified by an audio, video, or tactile alert through the CD. The information can then be sent from the requesting user to other net members after the user is granted the preferred priority for sending.

In one embodiment of the invention, each wireless network member establishes a forward link and a reverse link with one or more base stations 116 or satellite gateways 118, as the case may be. Base station 116 may be used to describe a communication channel from base station 116 or satellite gateway 118 to a CD. The satellite gateway 118 may be used to describe the communication channel from the CD to the base station or satellite gateway 118. Voice and/or data may be converted into data packets using the CD, such as appropriate for the particular distribution network 120 where communication to other users may occur. In one embodiment, the distribution network 120 is the Internet.

In one embodiment, a dedicated forward channel, i.e., a terrestrial communication system and a satellite communication system, is established in each communication system to broadcast information from each net member to the other net members over the dedicated channel. Each network member may receive communications from other network members on a dedicated channel. In another embodiment, a dedicated reverse link is established within each communication system to transmit information to the CM 110. In an embodiment, a combination of the above schemes may be used. For example, a scheme may involve establishing a dedicated forward broadcast channel but require wireless CDs to transmit information to the CM 110 on a dedicated reverse link assigned to each CD.

When a first net member wishes to transmit information to other members of the net, the first net member may request a transmission priority by pressing a push-to-talk button on his or her CD, which generates a request for formatting for transmission over the distribution network 120. In the case of CDs 102 and 104, the request may be sent over the air to one or more base stations 116. A Mobile Switching Center (MSC)122 that may include a well-known interworking function (IWF), Packet Data Serving Node (PDSN), or Packet Control Function (PCF) to process data packets that may exist between BS 116 and distributed network 120. For CD 106, the request is transmitted through satellite gateway 118. For CD 108, the request may be sent to modem bank 126 via Public Switched Telephone Network (PSTN) 124. The modem bank 126 receives the request and provides it to the distribution network 120. The NBS terminal 128 monitors traffic of the NBS system through its connection to the distribution network 120. Since the NBS terminal 128 is on the way to the distributed network 120, the geographic proximity to the network participants is not necessary.

If no other member currently holds a transmission priority, when the CM 110 receives a request to transmit a priority, the CM 110 may transmit a message to the requesting net member informing it that the transmission priority has been granted. Audio, video or other information from the first net member may then be transmitted to the other net members by transmitting the information to the CM 110 using one of the transmission paths described above. In one embodiment, the CM 110 then provides the information to other net members by copying the information and sending the copy to the other net members. If a single broadcast channel is used, the information need only be replicated once for each broadcast channel used.

In another embodiment, the CM 110 is included within the MSC 122 such that data packets from the supporting base stations are routed directly to the CM 110 and not to the distribution network 120. In this embodiment, the CM 110 remains connected to the distribution network 120 so that other communication systems and devices may participate in group communications. In another embodiment, the CM may be included in the PCF module of the PDSN or MSC.

In one embodiment, the CM 110 maintains one or more databases to manage information pertaining to personal web members and each defined web. For example, for each member of the network, the database may include information such as the user's name, account number, telephone number, or dialing number associated with the member's CD, the mobile station identification number assigned to the CD, the current member status within the network such as whether the member is participating in the network, priority codes to determine how to assign transmission priority, data telephone numbers associated with the CD, IP addresses associated with the CD, and instructions indicating that member authorization can communicate with those networks. Other relevant types of information associated with each net member may also be stored by the database.

In one embodiment, the CDs may form a connection of personal communication terminals to form a talk group or network. The CM may include a number of functions of hardware and software that may be configured in different ways to accommodate different applications. The CM may provide real-time, administrative, and authentication operations for the managed (NBS) network, push-to-talk (PTT) request arbitration, network member maintenance and distribution, and registry, call setup, and interruption of necessary communications, such as control of CDMA, system and network resources, and overall network status.

The NBS network may be within a stand-alone, dispersible cellular system or within a large multi-site configuration. In a large deployment scenario, multiple CMs may be geographically dispersed to form a single integrated system, each as a plug-in module to an existing cellular infrastructure. In this way, the new features introduced by the NBS network are available to cellular users without requiring modification to the existing cellular infrastructure.

The CM may maintain a table of defined NBS nets. In one embodiment, each net definition includes a net identifier, a membership table including telephone numbers or other identifying information, user preference information, and other types of administrative information. A net may be statically defined as either open or secure and does not allow for a transition between open and secure. Secure NBS networks typically use media encryption to provide authentication and to prevent eavesdropping. Media encryption for secure networks is implemented on an end-to-end basis, meaning that encryption and decryption may occur within the communication device. The CM may operate without knowledge of the security algorithm, key, or policy.

Fig. 2 illustrates an example NBS mesh 200 to show how a communication device 202 interacts with a CM 204. Multiple CMs may be deployed as needed to meet the expectations of a large-scale NBS network. In fig. 2, CD 202 has permission to transmit media to other net members. In this case, CD 202 is called a talker and transmits media over the channel. When CD 202 is designated as the interlocutor, the remaining network participants, CD 206 and CD 208, may not have permission to transmit media to the network. Accordingly, CD 206 and CD 208 are designated as listeners.

As described above, CDs 202, 206, and 208 connect to CM 204 using at least one channel. In one embodiment, the channels are divided into separate channels, including a Session Initiation Protocol (SIP) channel 210, an NBS media signaling channel 212, and a media traffic channel 214. Bandwidth allowance in the SIP channel 210 and NBS media signaling channel 212 may be used by the CDs 202, 206 and 208 at any time, regardless of whether they are designated as interlocutors or listeners. SIP is an internet engineering task force, IETF, defined application layer protocol that describes a control mechanism to establish, modify, and terminate media sessions operating over the Internet Protocol (IP). SIP provides a solution to call signaling problems typically applied to internet telephony by registering and locating users, mechanisms to define user capabilities and describe media parameters, and mechanisms to determine user availability, call setup, and call processing.

In one embodiment, the SIP channel 210 is used to start and terminate participation of CDs within the NBS network 100. A Session Description Protocol (SDP) signal may also be used within SIP channel 210. When participation of the CD within the NBS network is set, for example, by using SIP channel 210, real-time call control and signaling between the CD and CM may occur, for example, by using NBS media signaling channel 212. In one embodiment, the NBS media signaling channel 212 is used to handle push-to-talk requests and releases, arbitrate or floor control between conflicting requests, announce the start and end of information transfer, handle network dormancy, track endpoint connections, request and exchange network status and notify any error messages. The protocol of the NBS media signaling channel 212 minimizes the length of the most common messages and simplifies the task of interpreting answers and responses to requests and preserves future enhanced flexibility. The protocol of the NBS media signaling channel 212 also allows requests to be resent without adversely affecting the protocol state.

In an embodiment, signaling traffic on NBS media channel 212 includes call setup and control signaling, which may include session invite requests and acknowledgements, and media signaling, which may include real-time floor control requests and related asynchronous messages. Media traffic on media traffic channel 214 may include real-time point-to-multipoint voice and/or data broadcasts. Both message classes have unique functional characteristics. In addition, each CD may issue a Domain Name Service (DNS) client request to facilitate mapping a fully compliant DNS hostname to an internet address.

In an embodiment, NBS call setup and call control signaling is implemented according to SIP semantics. Although SIP may be transmitted using either the well-known User Datagram Protocol (UDP) or the Transmission Control Protocol (TCP), in one embodiment each CD uses SIP-based signaling functions using SIP. Also, each CM may desire to receive SIP signaling requests over UDP. Real-time signaling may occur through dynamic UDP/IP over the CM and each CD. Other signaling may occur, for example, using SIP over a fixed TCP/IP interface between the CM and the CD.

PTT latency

In one embodiment, when the packet data service is active, resources within the infrastructure, such as Base Transceiver Subsystem (BTS) { XE "BTS" }, Base Station Controller (BSC) { XE "BSC" }, Interworking (IWF) { XE "IWF" }, and radio links are actively allocated to the Mobile Station (MS). In the VoIP dispatch service based on IP { XE "IP" }, the packet data connection of each user remains active while an active conversation is in progress between the group participants. However, after a period of inactivity in the group communication, i.e., the "hang time," the user traffic channel may switch to a dormant state.

Switching to the dormant state conserves system capacity, reduces service costs and battery consumption and enables the user to receive incoming regular voice calls. For example, when a user is in an active packet data call, he is generally considered to be "busy" for other incoming voice calls. If the subscriber packet data call is dormant, the subscriber may be able to receive an incoming voice call. For these reasons, it is desirable to be able to switch a packet data call to a dormant state after a period of packet data inactivity.

When a packet data call is active, Radio Frequency (RF) energy may still be transmitted by the mobile station phone, even at lower levels, even if no data packets are being exchanged, in order to maintain synchronization and power control with the base station. These transmissions may cause significant power consumption by the phone. While in the dormant state, however, the phone may not implement any RF transmission. To conserve phone power and extend battery life, the hang time may be set to switch the phone to sleep mode for an extended period of time after no data transmission.

When the packet data service is active for all users, (PTT) { XE "PTT" } requests, which may be (IP) { XE "IP" } datagrams sent between the MS and the scheduling server, have a short latency. However, if the user channel previously switched to the dormant state, (PTT) { XE "PTT" } latency may be longer. State information associated with the packet data session, including the IP address of the mobile station, may be maintained after the packet data dormant period. However, state information related to layers below PPP, such as the physical traffic layer, may be released and/or de-allocated.

In some infrastructures, to wake up a dormant data connection, the traffic channel must be reallocated, resources must be reallocated and the radio connection protocol (RLP) { XE "Radio Link Protocol (RLP) } layer must be reinitialized. The consequence of this is that if the talk group has not been talking for some time, the PTT latency of the first talk spurt is generally longer than the latency of the successive talk spurt when the user presses his (PTT) { XE "PTT" } button to request the floor. This is generally infrequent, but may affect the use of the service and is therefore to be minimised.

{ XE "PTT" } in one embodiment, when a group communication device is in a sleep state, (PTT) { XE "PTT" } latency may be caused by:

1. talk channel assignment delay-the delay in assigning and initializing a traffic channel for the talker's telephone in response to the user pressing the talk button and the delay in initiating the IP floor request message by the scheduling application.

2. The talk request propagates delay-the time when the talk request message is passed to the dispatch server.

3. Arbitration latency-the time at which the scheduling server processes potentially multiple talk requests.

4. Wakeup message delay-the time for an IP message to travel from the scheduling server to the cellular infrastructure of the listener, e.g., PDSN.

5. Listener paging latency-the latency required to wait for the listener's phone to wake up and receive a page in the appropriate paging channel slot.

6. Listener channel assignment delay-the delay to assign and initialize the listener's telephone traffic channel.

Some of these delays have a greater impact on overall PTT latency than others. For example, the talker and listener channel allocation latencies and listener paging latencies are typically an order of magnitude greater than the other components, which together affect PTT latency performance.

To reduce PTT latency, in one embodiment, group call signaling, such as floor control requests, floor control responses, and dormancy wakeup messages, can be transmitted on some available common channels without waiting for the dedicated traffic channel to be re-established. Such common channels may be available at all times regardless of the state of the mobile station, and may not need to be requested and reassigned each time the user desires an initial group call. Thus, group call signaling may be exchanged even when the mobile stations are in a dormant state, which may provide a method of concurrently re-establishing dedicated traffic channels for the interlocutor and listener mobile stations.

In another embodiment, the calling mobile station may send a floor control request to the wireless infrastructure on some available reverse common channel, such as on a reverse access channel and a reverse enhanced access channel. The calling mobile station may also receive a response to the floor control request on some available forward common channel, such as the forward paging channel and the forward common control channel. In one embodiment, the dormant listener's mobile station may receive the dormant wake-up message on some available forward common channel, such as the forward paging channel and the forward common control channel.

Short data burst call signaling messages

In one embodiment, a substantial reduction in the actual total sleep wake-up time and (PTT) { XE "PTT" } latency perceived by the interlocutor may be achieved by using Short Data Burst (SDB) messages, such as those provided in "TIA/EIA/IS-2000 Standards for cdma2000 Spread Spectrum Systems," hereinafter referred to as "cdma 2000 Standards. In an embodiment, the SDB message may be sent on two dedicated physical channels, such as a forward Fundamental Channel (FCH) or a forward dedicated common control channel (F-DCCH), or a common physical channel, such as a reverse access channel (R-ACH), a reverse enhanced access channel (R-EACH), a forward common control channel (F-CCCH), or a Paging Channel (PCH). SDB messages may be conveyed by Radio Burst Protocol (RBP) { XE "Radio Burst Protocol (RBP) }, which maps the message to appropriate and available physical layer channels. Since SDB messages may carry any IP { XE "IP" } traffic and may be sent on common physical channels, SDB messages provide a mechanism to exchange group call signaling when the calling client's mobile station does not have a dedicated traffic channel.

Mobile station oriented call signaling messages

In one embodiment, the media signaling messages may carry IP datagrams on the reverse link or mobile station-oriented source link. A client mobile station may quickly signal a CM when a user requests the floor and there is no dedicated reverse traffic channel immediately available. Assuming that the client mobile station has released all dedicated traffic channels, the client mobile station may immediately forward the floor control request on the reverse common channel of the wireless infrastructure, which may forward the request to the CM. For example, either a reverse access channel or a reverse enhanced access channel may be used to send such messages when a dedicated reverse channel is not available. In one embodiment, the client mobile station may transmit the talk request message to the CM as an SDB message.

Figure 3 illustrates example call signaling for a floor control request procedure. A client Mobile Station (MS) may receive a request from a user desiring to initiate a group call. In an embodiment, the customer MS may be a PTT device. In an embodiment, the client MS may send a PTT talk request 302 on a reverse common channel, such as an access channel or enhanced access channel, before attempting to re-establish its dedicated traffic channel. In an embodiment, the client MS may send the PTT talk request 302 in an SDB message regardless of the channel used.

The customer MS may start re-establishing its dedicated traffic channel 304, for example by implementing "service option 33 re-origination". The client MS may also initiate Radio Link Protocol (RLP) synchronization 306. In an embodiment, the customer MS may re-establish its dedicated traffic channel and synchronize the RLP advantageously in parallel with sending the PTT talk request 302.

Thus, when the mobile station does not have an active dedicated traffic channel, the available reverse common channel and/or SDB feature is used to send a floor control request signal to the CM, reducing the total time required to wake up a participating mobile station. Although the talker client may not receive an acknowledgement of the authorized talk request before the talker's forward traffic channel is re-established, being able to quickly send to the CM to wake up the participating listener reduces the overall latency.

Referring to fig. 3, the wireless infrastructure may send a PTT floor control request 308 to a Packet Data Serving Node (PDSN) and then to the CM. In one embodiment, upon receiving floor control request 310, the CM may arbitrate the request, send a burst media signaling wake-up message (trigger) { XE "AYT wake-up request" } to a set of target participants (listeners), and/or trigger a re-establishment of participant (listener) traffic channels. If the CM grants PTT floor control, the CM may send a PTT floor grant 312 to the infrastructure, which may send a PTT floor grant 314 to the client MS. In one embodiment, if the customer dedicated traffic channel has not been re-established, the infrastructure may send a PTT floor grant to the customer MS on an available forward common channel, such as on a forward paging channel and a forward common control channel. In an embodiment, the infrastructure may send the PTT floor grant 314 to the client MS in SDB form regardless of the channel used.

In an embodiment, the CM may wait for the sleep response timer to expire before responding to the PTT floor control request. If the sleep response timer for the group is set to zero, the CM may respond immediately to the floor control { XE "PTT" } request. In one embodiment, if the client MS finishes re-establishing its traffic channel and RLP synchronization, the client MS may stream media 316, which may be buffered to the CM at the client MS.

Network-oriented call signaling messages

In one embodiment, upon receiving a floor control request, the CM may burst a media signaling wake-up message to a group of target participants (listeners) and trigger a re-establishment of the participant (listener) traffic channel. If the sleep response timer for the group is set to zero, the CM may respond immediately to the floor control { XE "PTT" } request. In one embodiment, if the interlocutor has begun to re-establish its traffic channel immediately upon sending a PTT { XE "PTT" } request, the caller's and listener's traffic channels may advantageously be re-established in parallel.

Fig. 4 illustrates example call signaling for a network initial sleep wake-up procedure. After the CM receives PTT floor control request 310 (fig. 3), the CM may send a wake-up trigger 402 to the target listener. The PSDN may determine whether a packet data session exists for the target mobile station and send a trigger packet to an appropriate infrastructure component, such as a base station. The infrastructure may page 406 to each individual target MS to begin re-establishing its dedicated traffic channel 408, e.g., to implement a "service option 33 re-origination". The target MS may also initiate radio connection protocol (RLP) synchronization 410. In one embodiment, the target MSs may re-establish their dedicated traffic channels and synchronize their RLPs to advantageously perform the same functions as some client MSs do in parallel.

In one embodiment, the CM may send a wake-up trigger 412 to the target MS after the target MS has completed re-establishing its dedicated traffic channel and has its RLP synchronized. The target MS may send a wake up reply 414 to the CM indicating that the target MS is ready to receive media. The CM may send a talker notification 416 to the client MS before streaming media 418 to the target MS, which media 418 may have been buffered within the CM.

In an embodiment, the infrastructure may send the wake-up trigger 412 to the target listener on some available common forward channels, such as a forward paging channel and a forward common control channel, while the traffic channel of the target listener has not yet been re-established. In one embodiment, the infrastructure may send the wake trigger 412 to the target listener in SDB form regardless of the channel used. If a PTT floor control request is sent as an SDB message on the talker reverse common channel and the dormant response timer for the target group is set to zero at the CM, the PTT latency actually at the talker client is reduced to the time required to send an SDB request message on the reverse link followed by an SDB response message on the forward link.

Network interface for call signaling messages

To determine what network-oriented particular traffic, e.g., SDB payloads, are sent to idle mobile stations that do not have dedicated traffic channels, some infrastructure policy or interface may be implemented that distinguishes this particular traffic from other traffic.

In a first embodiment, since SDB messages may carry limited user payload, IP datagrams may be filtered according to their size. If the destination is a mobile station without a dedicated traffic channel, IP datagrams smaller than a predetermined size limit can be sent as SDB messages. Such filtering may be used by group communication systems because application-floor request response messages are small, including, for example, 34 bits of IP header.

In a second embodiment, an infrastructure vendor may define an IP-based service to encapsulate IP traffic sent to a mobile station. An IP server with knowledge of the service may transmit a small IP, e.g., UDP, datagram, appropriately encapsulated with an IP header to the service for transmission to a mobile station suspected of not having a dedicated traffic channel. For example, the group communication system may use the service to indicate to the infrastructure that a talk request response message is sent to the requesting client MS in SDB form. Coordination of pending paging or service initiation requests with SDB traffic is important for fast and reliable user traffic delivery.

In a third embodiment, the IP server may transmit a specific IP, e.g., UDP, datagram with IP headers for transmission to a mobile station suspected of not having a dedicated traffic channel. The IP server may mark the IP datagram, for example by indicating a specific value in the IP header, instructing the infrastructure to send the IP datagram to the client MS. For example, the group communication system may use the service to indicate to the infrastructure that a talk request response message is sent to the requesting client MS in SDB form. In a third embodiment, UDP or TCP port ranges may be reserved for sending specific IP datagrams, e.g., SDB messages.

Mobile station originated service initiation and paging

In one embodiment, as discussed above in connection with fig. 3, a conversational Mobile Station (MS) may send a floor control request to the CM, possibly in the form of an SDB, and then immediately send a service initiation request 304 to the wireless, e.g., CDMA, infrastructure to quickly re-establish its traffic channel. However, if the sleep response timer is set to a small value, the CM may respond quickly to floor control request 310 and transmit response 312 back to the talker MS. If the response arrives at the infrastructure at an early stage in the service initiation transaction 304, the infrastructure records that the talker MS does not have any active traffic channel and attempts to page the response to the talker MS. However, the paging action may abandon the service initiation process already in progress. In one embodiment, the interlocutor MS may respond to the page, guarantee that a floor control response message is sent to the interlocutor, and request service initiation again, but there is inevitably a delay in re-establishing the interlocutor traffic channel as a result of abandoning the initial service initiation attempt.

In a first embodiment, to avoid race conditions between the service initiation process and paging, the CM may be configured not to respond immediately to floor control request 310. Accordingly, a sleep response timer, e.g., within the CM, may be adjusted such that the CM transmits a response 312 to the interlocutor MS upon completion of the service initiation process 304.

In a second embodiment, a PDSN receiving a CM originated response 312 is coordinated with a mobile station switching center (MSC) responding to a talker service origination request. I.e., if the PDSN determines that the packet data service initiation process for the interlocutor MS is in progress when the CM originated response 312 arrives at the infrastructure, the PDSN may cache the response and send it on the forward traffic channel of the interlocutor MS once the service initiation process is complete. Alternatively, if the service initiation procedure is still in progress, the MSC may send the response as an SDB message to the conversational MS.

In a third embodiment, the interlocutor MS may avoid a race condition by sending the service initiation request 304 after the interlocutor MS has received a response to the floor control request 302. In one embodiment, since the talker MS has no active dedicated traffic channel, the CM may send a response to the talker MS on some available forward common channels, such as a forward paging channel and a forward common control channel. In one embodiment, the CM may send the response to the interlocutor MS in SDB form. The talker MS may rely on the CM generated floor control response 312 to trigger reactivation of its traffic channel in the same manner as the listener's mobile station's CM triggered a wake-up request sent by the traffic channel reactivation. Race conditions are avoided because of the avoidance of the possibility of a mobile-originated service origination occurring simultaneously with a mobile station's network-originated page.

Caching network originated packet data triggers

IP datagrams arriving at the wireless, e.g., CDMA, infrastructure that include the wake-up trigger 402 and are sent to a listener mobile station that does not have a dedicated traffic channel may typically be lost by the network or the wireless infrastructure. In one embodiment, the wake-up trigger 402 sent to the listener's mobile station is continually re-sent according to a defined schedule until the listener's response or group wake-up timer expires. For example, the wake-up trigger 402 may be retransmitted every 500 ms. However, retransmitting the wake-up trigger 402 at this rate may cause a maximum delay of up to 500ms, or an average delay of 250ms, which is the time from the listener's traffic channel re-establishment to the listener's next wake-up trigger arrival at the infrastructure.

In an embodiment, an infrastructure or other entity within the network may cache the wake trigger 402 sent by the CM and send it to the target MS once the target MS has re-established its traffic channel. This eliminates the need for the CM to resend the wake-up request 402 and reduces the total sleep wake-up time. E.g., the cache wakeup trigger 402, delays of up to 500ms may be omitted from the total sleep wakeup time as opposed to retransmitting at a rate of 500 ms.

Media buffering

In one embodiment, the user may be allowed to start a conversation after the user requests floor control by buffering the media before reestablishing the dedicated channel between the client and the listener. By buffering the talker's talk, the system allows the talker to start talking before the listener's traffic channel is fully established. This allows the interlocutor to start talking earlier, significantly reducing his PTT { XE "PTT" } latency. Since listeners do not experience PTT latency, they are not affected, i.e., PTT latency is shifted from the interlocutor to other parts of the system. The interlocutor may just wait for the time from when the listener receives a response to his first talk burst, but as mentioned above, he knows that he takes longer for the response of his first talk burst than for his next talk burst between active conversations. Buffering the first talk burst of the interlocutor can be done at the CM side or at the client MS side.

CM buffering

In one embodiment, the CM may buffer the first talk burst of the interlocutor. After the user presses the PPT { XE "PTT" } button and the user's traffic channel is re-established, he may be allowed to communicate with the CM. At this point, the CM buffers the talker for future delivery to the target listener, since the listener traffic channel has not yet been established. CM buffering may reduce the apparent PTT { XE "PTT" } latency experienced by a talker to the time used to establish the conversational traffic channel. Fig. 5 illustrates CM buffering according to an embodiment.

Client side buffering

In an embodiment where a shorter apparent latency is desired, the interlocutor may be allowed to begin speaking before his traffic channel is re-established. Since the client MS has not communicated with the CM yet, a signal to the interlocutor to start talking is formed by the client MS. If the talker is allowed to speak before the conversational traffic channel is re-established, the client MS may buffer the utterance. Since communication with the CM has not been established, the permission for the dialog is given "optimistically". In an embodiment, the CM buffering and the client-side buffering may operate simultaneously. Client-side buffering may allow explicit PTT { XE "PTT" } latency to be small.

For CM buffering, the total latency does not change. The user may experience the same delay in receiving the response from the listener, but the apparent PTT { XE "PTT" } latency of the interlocutor may be made small.

In one embodiment, the client MS may buffer the media to control the apparent PTT latency experienced by the user. The combination of mobile-oriented SDB and client-side media buffering may reduce the latency associated with re-establishing an active traffic channel.

Quick paging channel

In one embodiment, the CM may defer answering to the talker PTT { XE "PTT" } request until the group's wakeup timer expires or all listener clients have responded to a network-originated trigger to establish their respective traffic channels. The CM may wait until all listeners have been paged before allowing the interlocutors to stream media within the group. The longer the group listener responds to the page, the longer the PTT latency is perceived by the talker.

In one embodiment, during sleep wakeup, each listener client individually receives a series of wakeup triggers sent by the CM, triggering one or more pages to each mobile station upon arrival at the infrastructure, e.g., CDMA. Upon receiving the page, each mobile station may re-establish the traffic channel, receive the next wake-up request sent to it, and respond to the CM with a wake-up request reply { XE "IAH reply" }. A significant portion of the time required for the listener handset to respond to the application layer "ping" takes the infrastructure time to wait for the mobile station to be properly paged.

To conserve battery life, the mobile station may not need to constantly monitor each of the defined, e.g., 2048, time slots { XE "Forward Paging Channel (F-PCH)" } within the Paging Channel when the mobile station is in an idle state. Instead, the mobile station may monitor either a forward common control channel (F-CCCH) or a forward paging channel (F-PCH), depending on the capabilities of the mobile station. In addition, the mobile station may monitor the paging slot based on the slot cycle index.

In one embodiment, to conserve battery, the mobile station may operate in a "slotted paging" mode. In this mode, the mobile station periodically wakes up for a short period of time to listen for pages sent by the Base Station (BS). The BS may know when the mobile station will listen and may send a page to a particular mobile station during a particular paging slot.

In one embodiment, the period during which the mobile station wakes up to listen to the paging channel is controlled by a parameter called Slot Cycle Index (SCI) { XE "Slot Cycle Index (SCI) }. The larger the SCI, the longer the time between slots the mobile station wakes up to listen to the paging channel. The large slot cycle value increases phone standby time due to telephoning for longer periods of time in sleep mode and also increases the time that the BS may need to wait before it can page the phone.

The time that the BS may need to suspend paging of the phone varies during the full slot cycle starting at zero if the slot of the phone is just at the time the BS needs to page it to ending if the phone slot is just at the time the BS needs to page the phone. On average, the delay due to waiting for a phone slot to arrive is half a slot cycle time. The shorter the slot cycle used by the mobile station, the faster the listener may be paged by the infrastructure. However, a shorter slot cycle may mean that the battery is drained more quickly.

In one embodiment, the Forward Quick paging channel (F-QPCH) { XE "Forward Quick paging channel (F-QPCH) } may be used to allow the mobile station to determine when a pending page arrives in a more power efficient manner without requiring the mobile station to monitor the paging channel itself. A mobile station capable of monitoring the F-QPCH may wake up every predetermined number of slots to extract a one-bit indication value in, for example, an 80ms slot on the paging channel. If the extracted bit is not set, there is no pending page on the paging channel and the mobile station sleeps for another slot cycle. If the extracted bit is set, a page to the mobile station is pending and the mobile station may rearrange itself to wake up and monitor the paging channel at the next appropriate paging channel slot.

The modulation used by the F-QPCH enables the mobile station to monitor the F-QPCH more efficiently than monitoring the paging channel. This enables the mobile station to operate efficiently in a very short slot cycle in a power saving manner. One benefit of using the F-QPCH is to provide that the mobile station detects and responds to general paging messages from the infrastructure and thus wakes up request messages from the CM at a faster slot cycle than the same battery consumption rate. This in turn becomes the ability to minimize the component of the delay that directly causes PTT latency and the total sleep wakeup time-the time needed to re-establish the listener's traffic channel.

Time slot timer

In one embodiment, the mobile station may operate in a non-slotted paging mode in conjunction with a "slotted timer". When activated, the slot timer requires the mobile station to monitor the paging channel in non-slotted mode when releasing its dedicated channel and entering an idle mode time period defined by the slot timer. The value of the timer may be configured at the base station. This feature allows the infrastructure to instruct mobile stations in idle mode to monitor the paging channel for slots, e.g., every 80ms, and provide the infrastructure with a method to page the mobile station in any slot. One advantage of using non-slotted mode, as in the case of the quick paging channel feature alone, is that it provides the mobile station with a way to detect and respond quickly to pages at the same rate of battery consumption, thus reducing the time required to re-establish the listener's traffic channel during sleep wakeup.

Without the quick paging channel feature, the extended use of non-slotted monitoring can be very battery consuming. However, using the quick paging channel together with the non-slotted mode provides a method of almost immediately paging the mobile station-within one or two slot periods, e.g., 80 to 160 ms.

The non-slotted mode may be considered one of two sleep intermediate phases available to the mobile station. When operating in non-slotted mode, the mobile station may be considered technically dormant because it has no dedicated physical channel. However, in this mode, the mobile station may be paged immediately in any time slot, thus avoiding the paging delay associated with network-initiated reactivation.

Control hold mode

In one embodiment, the mobile station may operate under a packet data standard that provides an additional dormant/idle state in which the mobile station and infrastructure maintain a PPP layer state associated with the mobile station while allowing one of the two endpoints to release dedicated traffic signals and other resources associated with the mobile station's packet data service option call. Both the mobile station or the infrastructure may transition the state of the packet data call from the dormant/idle state to the active state by re-establishing the traffic channel and re-negotiating the RLP. The time required to re-establish the traffic channel may depend on the mobile station or infrastructure to begin re-establishment. In both cases, however, the delay is comparable to the need to initiate a new call on the system, since all system resources may need to be requested and allocated to the mobile station.

In one embodiment, the mobile station may operate in a "control hold" mode as an intermediate position between active and idle modes. In the control reservation mode, the dedicated traffic channel associated with the mobile station may be released and the mobile station's reverse pilot may operate in a "gated" mode. In an embodiment, dedicated common control channel and/or RLP states may also be maintained. In essence, the control reservation mode provides a semi-dormant state in which most of the system resources may be reserved for allocation, but the average reverse link transmission power is reduced to the gated pilot to reduce the impact on system capacity. Fig. 7 shows an example arrangement of radio modes.

In one embodiment, the mobile station may transition from the active mode to the control reservation mode by either sending a resource release request message or a resource release request short message. The mobile station may transition from the control reservation mode to the active mode by sending either a resource request message or a resource request short message. These messages may be transmitted over a dedicated control channel and the short messages may be sent using shorter frames, e.g., 5ms, allowing for fast switching between being in and not being in the control reservation mode. An advantage of the control reservation mode over the conventional idle mode or sleep/idle mode as described above is that the transition from the control reservation mode to the active mode is relatively fast.

In one embodiment, upon receiving a signal from the CM indicating that the subscriber group has transitioned to the group dormant state, the client mobile station may begin transitioning to the control hold mode and further to the idle mode after an additional hold inactive state. Thus, controlling the reservation mode provides a mechanism to greatly reduce the time required to re-establish a dedicated traffic channel once a user presses PTT or a wake-up request trigger is received at the infrastructure.

Stored service configuration

In one embodiment, the infrastructure may provide the ability to buffer or store service configuration states at the mobile station and infrastructure when transitioning to idle mode. When returning to active mode and re-establishing the traffic channel, the mobile station may indicate in the origination message or page response message that it has buffered or stored the service configuration for the call. The mobile station may also include a Cyclic Redundancy Check (CRC) within the origination or paging message, which may be calculated over the entire length of the service configuration. If the base station also caches the service configuration, the base station may use the received CRC to confirm that its service configuration conforms to the mobile station's stored service configuration and, if so, the BS may indicate within its "service connect message" that the mobile station may use the previously stored service configuration.

In an embodiment, the use of the packet data service option may not require a change in service configuration when transitioning out of idle mode, and thus the use of the stored service configuration may result in a substantial reduction in the time required to cause re-establishment of dedicated traffic channel resources. Thus, the storage service configuration feature provides a mechanism to greatly reduce PTT latency by reducing the time required to re-establish traffic channels that may carry PTT signaling and related media, thereby achieving a significant enhancement to idle mode.

In an embodiment, the following may be implemented for the client MS to go from active mode to idle mode:

1. the group is active and the mobile station has a dedicated traffic channel.

2. An application layer group dormancy notification is received on a forward traffic channel of the mobile station after an inactivity period that exceeds the group hang time timer.

3. The mobile station goes to the control hold mode and caches the state of its service configuration. Similarly, the base station of the customer also caches the service configuration state.

4. After the inactive period, the mobile station releases its dedicated channel and transitions to idle mode. The mobile station begins monitoring the quick paging channel and may enter a non-slotted mode if there is an infrastructure indication. If the period of inactivity is relatively short-either due to a local user pressing PTT or due to network initiated packet data traffic from another group of participants-the mobile station may not reach idle mode before transitioning back to active mode. In this case, the mobile station will quickly switch back to the active mode since it has reserved its dedicated channel.

In one embodiment, the sleep wake event is implemented as follows:

1. the group is dormant and all mobile stations are in a state where no dedicated physical channel is idle. The mobile station monitors the quick paging channel.

2. In response to the user pressing down to talk, the mobile station of the interlocutor sends a signal to the CM on some available reverse common channel along with an application layer talk request message, which may be in the form of a short data burst. The interlocutor's mobile station may buffer the user media forward from that point on.

3. The mobile station of the interlocutor sends an "origination message" to the infrastructure to re-establish its traffic channel. It may indicate within its request that it has cached the service configuration and may include a CRC on the configuration data. This begins the process of re-establishing the mobile station traffic channel of the interlocutor.

The CM receives the floor request and decides through an arbitration process whether to grant the request and sends a floor request response message to the talker. The CM also starts bursting a series of wake-up requests to all participants.

5. Upon receiving each wake-up request, the infrastructure pages each listener's mobile station by first determining the next appropriate slot to page the listener's mobile station, and then signals through the F-QPCH that a page is pending for the listener's mobile station on the paging channel prior to that slot.

6. Upon receiving an indication on the F-QPCH that paging is pending, each listener mobile station monitors the paging channel for pages.

7. Upon receiving a page on the paging channel, each listener mobile station responds to the page indicating in its page response that it has cached the service configuration and may include a CRC on the configuration data. This begins the process of re-establishing the traffic channel for each listener.

8. After the talker traffic channel is re-established. A next talk request response from the CM is received at the interlocutor. The interlocutor begins streaming media to the CM.

9. After re-establishing each listener traffic channel, the next wake-up request sent by the CM is received at the listener. The listener replies with a wake-up response message.

10. Once all listeners respond or the group wakeup timer times out, the CM starts to stream media to the group.

Thus, the method and apparatus for reducing latency in a group communication network disclosed herein greatly reduces the actual total dormancy wakeup time and PTT { XE "PTT" } by exchanging group call signaling even if the mobile station is dormant and no traffic channel is active. The method and apparatus provide for exchanging group call signaling using Short Data Burst (SDB) message signaling. The method and apparatus provide for advantageously parallel re-establishing dedicated traffic channels for a talker mobile and a dormant listener mobile.

In another embodiment, sleep wake-up latency in a group communication network may be reduced by caching network-originated wake-up triggers to target listeners and sending the wake-up trigger to the target mobile station once the target mobile station has re-established its traffic channel.

In another embodiment, simultaneous service initiation and paging operations of a mobile station within a group communication network are avoided by transmitting a response to a floor control request after completion of a service initiation procedure. In an embodiment, the response to the floor control request may be in the form of SDB if the service initiation procedure is not complete. In another embodiment, the service initiation procedure of the source communication device is started after transmitting the response to the source communication system.

Claims (8)

1. A method of reducing latency in a group communication network, the method comprising:

receiving a sleep-wake trigger directed to a target communication device;

triggering the target communication device to re-establish its traffic channel; and stores the sleep-wake trigger for later transmission to the target communication device.

2. The method of claim 1, further comprising:

the stored sleep-wake trigger is sent to the target mobile station once the target mobile station has re-established the traffic channel.

3. A computer-readable medium embodying a method of reducing latency in a group communication network, the method comprising:

receiving a sleep-wake trigger directed to a target communication device;

triggering the target communication device to re-establish its traffic channel; and stores the sleep-wake trigger for later transmission to the target communication device.

4. The computer-readable medium of claim 3, wherein the method further comprises:

the stored sleep-wake trigger is sent to the target mobile station once the target mobile station has re-established the traffic channel.

5. An apparatus for reducing latency in a group communication network, comprising:

means for receiving a sleep-wake trigger directed to a target communication device;

means for triggering the target communication device to re-establish its traffic channel; and

means for storing the sleep-wake trigger for later transmission to the target communication device.

6. The apparatus of claim 1, further comprising:

means for transmitting the stored sleep-wake trigger to the target mobile station once the target mobile station has re-established the traffic channel.

7. An apparatus for reducing latency in a group communication network, the communication device comprising:

a memory cell;

a receiver;

a transmitter; and

a processor communicatively coupled to the memory unit, the receiver, and the transmitter, the processor capable of:

receiving a sleep-wake trigger directed to a target communication device;

triggering the target communication device to re-establish its traffic channel; and stores the sleep-wake trigger for later transmission to the target communication device.

8. The apparatus of claim 7, wherein the processor is further capable of:

the stored sleep-wake trigger is sent to the target mobile station once the target mobile station has re-established the traffic channel.